Testing ACTi E97 10MP CameraBy: Ethan Ace, Published on Jan 13, 2014
$200 for a 10MP camera? Too good to be true or the next big thing to stand out from the commoditizing HD market?
After years going sideways, ACTi is pushing hard with an ambitious, broad line of low cost products. Their recent release of the 10MP E97 minidome is perhaps the best example, drawing a lot of attention for its $200 price.
Non proprietary, H.264, 10MP cameras are pretty rare. Arecont has the widest offerings here. And, interestingly, Arecont likes to brag that camera decisions are all about getting the most pixels per dollar. Indeed, if that is the case, this new ACTi blows Arecont away.
Low cost and higher resolution are clearly two of the biggest needs in surveillance, so we bought an E97 and matched it up against the Arecont 10MP (which we tested here) as well as the Axis M30 1080p (tested here).
Here's what they look like in our test setup (from left to right, ACTi E97, Arecont AV10115, Samsung SNB-5004, and Axis M3005):
Our goal was to determine:
- Whose 10MP image quality is better - the ACTi or Arecont? Both in day and night scenes
- How much of a difference did the 10MP ACTi and Arecont do against the 1080p Axis? Both in day and night scenes
- What type of bandwidth premium did the 10MPs require compared to the 1080p?
- Overall, what benefit, if any, does a $200 10MP deliver compared to current offerings?
- Performance in full light at closer range (~25' to target, ~60' horizontal FOV) slightly better than the 10 MP Arecont AV10115, with more details of subject and chart.
- Full light performance at longer range (~40' to target, ~100' HFOV) moderately less clear than Arecont AV10115, with chart much less visible, due to poorer light handling and lens quality.
- Low light performance better than Arecont AV10115, with detection possible, but no details provided.
- Substantial issues with factory focus, forcing us to partially disassemble the camera to refocus.
- Bandwidth consumption slightly lower than Arecont AV10115 in full light (though much higher in low light, since the AV10115 produced an essentially black image).
- Bandwidth, compared to 1080p cameras, at the same frame rate, was 8x greater for the 10MP ACTi camera, a massive increase.
- DRF increased in low light from 30 (fairly typical compression) to 41 or higher, drastically reducing quality. This occurred no matter which streaming method was used, VBR or CBR.
ACTi E97: ~$200 USD Online [Note: this is listed as promotional pricing through March 20, 2013. From what we have seen pricing might ultimately be in $300 to $400 range.]
- Arecont 10115AI: ~$650 Online (Does not include lens)
- Axis M3005: ~$350 Online
- Samsung SNB-5004 [link no longer available]: ~$390 Online (Does not include lens)
- The Arecont AV10115 as a competitive 10MP camera.
- The Axis M3005V, in order to show differences between 1080p and 10MP across a similarly wide field of view (101 vs. 118).
- The SNB-5004 as a comparison against a low light optimized camera.
- ACTi E97: A1D-500-V6.05.16-AC
- Arecont 10115AI: 65221
- Axis M3005: 126.96.36.199
- Samsung SNB-5004: 1.11_131011
Here are our key findings from the test:
The E97 offers distinct advantages in wide, evenly lit fields of view, especially taking into account its much lower price than competitive 10MP options. However, at longer ranges, some clarity is lost, reducing its usefulness for identification. If low light is a key concern, users would be better off using 1080p resolution or lower, as only detection is possible, at most, using the E97. Additionally, be prepared for a sizeable increase in storage costs, given the ~8x greater bandwidth consumption.
Alternatively, ACTi is releasing other 10MP cameras in 2014 that include integrated IR and varifocal lenses, which may be a better option for those wanting a professional camera with higher resolution and better low light performance.
Factory Focus Issues
We found substantial focus issues with the E97 out of the box, despite claims that it ships pre-focused from the factory. Refocusing may be accomplished by unscrewing a locking ring and simply turning the M12 lens, though this easily adds ~10-15 minutes. The image below shows the improvement in performance between factory focus and refocusing for testing:
We recommend to immediately check / refocus during initial setup rather than taking a risk that the camera was shipped out of focus.
In this video we review the physical form factor of the camera and key features:
Web Interface Overview
This video reviews the web interface of the E97:
Video Samples Download
Below are video clips of the ACTi E97 in use in both full light and dark scenes at the camera's max 6 Frames Per Second.
We shot the 10MP ACTi E97 in our conference room and office scenes in both full and low light environments. We tested the E97 against multiple cameras:
This image shows the FOV used in this test:
We placed our subject about ~25' from the cameras in order to test differences in image quality. At this light and PPF level, the ACTi E97 performs well, with more lines legible than the Arecont AV10115 (down to line 6 vs. line 5), and clearer details of our subject.
In low light, the E97 can just barely see our test chart and subject, while the Arecont 10MP provides no usable image. The low light optimized Samsung SNB-5004 easily outperforms all the other cameras in the test.
We moved our test setup to a larger area, giving us a 100' horizontal FOV in order to see differences in performance at longer range.
This image shows the FOV used in this scene:
With our subject at this distance, the E97's performance lags behind the AV10115, with only the first line of the chart fully legible, though subject details are roughly the same. The M3005 is easily outperformed by the E97 at this range.
At this range, the E97 is unable to detect our subject. The M3005 and SNB-5004 are able both able to detect the chart at this range, but not the subject.
In our tests, no matter which streaming method we chose (CBR or VBR), bandwidth stayed steady, and quantization increased in low light (from ~30 to ~41 using "high" quality). Based on this, it seems that the E97's VBR implementation is broken, and CBR is used no matter what is selected.
Using all cameras' maximum framerates, the The E97's bandwidth usage was lower than both the Arecont AV10115 and Axis M3005 in full light, though it had the highest consumption in low light. However, standardizing all cameras at 6 FPS (the E97's maximum), consumption of the M3005 drops significantly to about .5 Mb/s, the lowest bandwidth in the test.
Users should note this huge increase in bandwidth at similar framerates when weighing tradeoffs of 10MP, as this will result in much higher storage needs, as well.
All cameras were tested using default settings, except for exposure, which was standardized at 1/30s. The following firmware versions were used in this test: